Artificial Intelligence Borrowing from the Software as a Service Playbook

Portfolio Manager Denny Fish explains why the industry- and function-specific nature of AI applications is conducive for multiple specialized AI developers to thrive.

Key Takeaways

  • We expect the AI landscape to evolve similarly to the way SaaS did over the past decade, as a wide range of stocks generated attractive returns.
  • As with SaaS, the highly specialized nature of AI applications lends itself to a relatively fragmented industry where niche companies can thrive in contrast to a winner-take-all ethos.
  • While the opportunity is large, the high costs and complexity of developing AI applications may lead to higher failure rates compared to SaaS, thus raising the importance of deep, fundamental research to identify potential winners.

Within the equities market, technology stocks have the tendency to dominate investors’ attention. This seems justified given the sector’s large market capitalization, its market-leading returns over the past decade and the role tech plays in writing the operating system for a global digital economy. Yet the headline grabbers tend to be the large Internet platforms and hardware producers whose industries are governed by a winner-take-all mentality.

Sometimes overlooked are some of the other themes that are reshaping the sector, including the Internet of Things, the transition from on-premises servers to the cloud, the rise of mobile computing and the increased level of connectivity needed to enable each of these developments. Unlike the Internet, some of these industries are highly fragmented as they cater to specialized business verticals and, thus, provide an environment where many companies can thrive. This is what has occurred with Software as a Service (SaaS). Rather than being dominated by a few mega-cap players, the SaaS industry is one where there has been a multitude of smaller winners, each providing value and efficiencies within a unique industry or function.

The themes propelling the tech sector are both complementary and iterative; the next stage is reliant on earlier developments and at the same time reinforces their position. Within the next iteration, we expect artificial intelligence (AI) to play a prominent role. We also believe that, given the highly fragmented nature of the landscape and the breadth of opportunities, the AI industry will follow a path similar to the one traveled by SaaS.

A Generational Transition

Over the lifetime of the industry, computing has transitioned from mainframes, to servers and personal computers and, more recently, to the cloud. This last step catalyzed the SaaS industry, enabling software developers to shift from a licensed to a subscription business model. This created an environment where a range of SaaS companies could thrive by offering highly specialized business applications.  While there has been some consolidation, the SaaS industry remains one where several players continue to leverage their unique technologies to add value to customers and their subscription-based offerings to generate steady revenues.

Building on Past Successes

While many tech investors continue to look for the next big idea with the most durable competitive advantages that can be leveraged to grab the highest market share, we think potential opportunities may lie in smaller, specialized firms, among them AI developers.

By its nature, AI lends itself to addressing complex, industry-specific tasks and achieving unique functional objectives. Already we are seeing AI deployed in a range of industries including agriculture, health care and finance. Management teams are recognizing the power of AI for finding solutions in complex environments, from monitoring soil conditions and weather on farms to using machine vision to perform quality control on a factory floor in real time.

From a functional standpoint, AI is being used in both front-office and back-office applications. In the former, it’s seen as a powerful tool to drive sales, limit churn and increase the customer experience. We are likely not far from a time when AI-powered bots can deliver a more satisfying customer interaction or at least guide staff by providing real-time feedback. Within the back office, verticals such as financial processing will likely be streamlined by function-specific AI tools to eliminate inefficiencies and reduce costs. In each of these examples, the targeted task of a particular AI-enabled application is likely narrow, thus allowing many providers to flourish.

Early Days … but Accessible

While AI’s ability to address unique tasks is similar to that of SaaS, there are differences. Developing an effective AI tool is more complex than it is for SaaS applications. That may translate into higher costs and higher failure rates. Yet this point reinforces why investors should consider a wide universe of companies, knowing that some will be successful, some won’t, and others may join forces or be acquired by larger companies that offer complementary services.

We are early days still in what may ultimately be called the AI Revolution. Many of the most innovative developers of AI remain private, thus limiting their access to investors. Yet other channels exist through which investors can gain access to this powerful theme. Large Internet platforms, for example, can offer the infrastructure and algorithms needed by independent AI companies. Developers may have access to data sets and industry expertise, but lack the firepower necessary to develop an effective AI tool. And as these technologies are iterative, some of this past decade’s cloud and SaaS winners can possibly provide data and expertise to new AI companies or be potential suiters should their markets and services be aligned.

In a broader sense, just like with SaaS adoption, we expect AI to be disruptive across a range of industries. Investors will need to watch how individual companies embrace – or ignore – the power of AI to help determine who will be the disruptor and who will be the disrupted.

ABANDON YOUR DOUBTS,

NOT YOUR GOALS

ABANDON YOUR DOUBTS,